system surveillance library/20051809.pdf · 2011. 9. 16. · {-j aspects monitored system...

("~----------_._-~

MODULE 8•••••••llm

SYSTEM SURVEILLANCE

2

MODULE 8 OBJECTIVE•••••••llm

II To understand the need for and the benefitsderived from a systematic review ofpla::tequipment and performance.

• To appreciate the importance of applyinglessons learned to improve future plantperformance.

II To recognize the importance of observingthe performance of other plants to avoidfuture potential difficulties.

System Surveillance

TOPICS•••••••llm

• Reason(s) for surveillance

• Aspects surveyed

• Methodology for effective surveys

• Example ofvarious cases

TOPICS•••••••lIl11m

€T~V

• Assessment of findings

• Link to future performance

• Effectiveness of surveillance programs

IL...- ---.-J

(

4 System Surveillance

REASONS FORSURVEILLANCE

_ ••••••IID~

• Good business sense, ensure systems andequipment work as required.

• Legal requirements, condition of operatinglicense.

II Input for business planning for future workprograms

• Building a data bank of the performancehistory of equipment and systems.

6

ASPECTS SURVEYED•••••••llm

• Overall plant perfonnance- Station capacity factor

- Scheduled incapability

- Forced incapability

- comparison between utilities: PWR'S ; BWR'S

---------~-------

System Surveillance

{-j ASPECTS MONITORED

System surveillance is a very extensive topic and the success of a plant restslargely with what is surveyed and more importantly what is done with theresults. Equally important is what is NOT surveyed and what impact this mayhave on plant operation while problems may go on undetected.

The surveillance is multi- dimensional and should address every area ofplantoperation ie. technical, maintenance, operations, load at various aspects ofthecritical systems, and examine the performance ofthe systems, equipment andcomponents.

As an example typical causes of incapability are measure and reported egoDiagram, OHN Causes OfIncapability.

The potential benefits of good surveillance are shown on 'Benefits InCapacity Factor'.With the appropriate effort and work programs in place about 20% loss isavoidable.

Some of the majcr gains in capacity factor can be achieved by carefulmonitoring and good measuremerlt of the main thermal cycle. Loss ofMWoutput can readily be traced to fouling of the main condensers, air illle!lkage,reheater and reheater drains problem and fouling of the steam generators.

""""

00

OHNCAUSES OF INCAPABILITY

a'NCE IN 'l!RVtCE TO lin tIN

1.1

6.0 8.0 10.0 12.0

PlANlIIED IfWll FORCED_

I

0.7

t. ......................,

13

0.4

0.4

0.1

U

0.1O.t

'.2O.S0.2

0.'O.t0.2

0.2

•••U

••1.1U

, , .

I!~

•••...OA

••• d ••••••• · ..... " ""' .."""..... ".,,,,,,,.,,.....

U

U

1.2

O.S

U0.2

1.1.......................................'.1.......................0.1

c.

10.6 8.0 8.0 4.0 2.0 0.0 2.0 .uGROSS UNIT ICbF (%)

12.0

,,,",Chan_.__......PHTT_

a-y•.-.M_,Ml__v

c-.-mFual Han<IIlng

_ConIr6I1

F_EII.lftCII

....MVC~1ng

I'lIT~'"

COIMnIlonll Aux.

Nue"'Awl.-..._lUIH.......

~ll"CCAdItI_

. .

ASPECTS SURVEYED•••••••lllm

• Overall plant perfonnance- Station capacity factor

- Scheduled incapability

- Forced incapability

- comparison between utilities: PWR'S ; BWR'S

L _


SYSTEMS SURVEYED•••••••llm

• Service systems- Capacity of D.C. batteries

- Instrument air tank capacities

-- Chiller system performance

12

SYSTEMS SURVEYED•••••••llm

• Key process systems control systems- Reactor regulating system

- Control computer reliability

- Liquid zone system

- Heat transport pressure inventory and control

- Boiler level control

- Boiler feed water system

System Surveillance

(

14

ASPECTS MONITORED•••••••lllIm

• Major economic influences- Thennal perfonnance of power" train.

» Boilers

» Steam system

» Turbine

» Condenser

» Feedwaters

» Electrical output

System Swveillance

(

r-'. _.'./

ASPECTS MONITORED•••••••llIm

• Key items within systems- Electrical relays

- Ground faults

- Instrument calibration drift

- Reference leg plugging

• System chemistry- H2 inPHT

- 02 in feedwater

16

METHODOLOGY OFEFFECTIVE SURVEYS

•••••••IIDD• Much work has been done in the nuclear

industry in efforts to optimize thesurveillance process.

• It is generally accepted that organizationshave to be selective in the choice of thesystems which will provide the greatest payback.

• The initial step of selecting which system tomonitor and \vhy is crucial to the success ofan optimuln program.

System Surveillance

(


•••••••II!IO~

• Once the systems have been selected forsurvey, the following general review shouldbe carried out on each system:- Define perfoTIllaI!ce goals / indicators for

system

- Define the importance of system function andcomponents

_J

".;-?:--';

z~ _f

18


•••••••liD~

- Define system motoring requirements

- Identify the data required

- Identify actions required

- System monitoring documentation

System Surveillance

METHODOLOGY FOREFFECTIVE SURVEYS

•••••••••ou• n carrylllg out t e system reVIew, It IS VitaI

that the various views of system strengths andweakness be identified.

• A team approach is required and should bemade up ofknowledgeable stafffrom- Technical

- Operation

- I & C / electrical maintenance

- Mechanical maintenance

- Manufacturer - where available

~GuidelinC$ (or System Monitoring by System Enajneers"(a discussion of an EPRl-PSE work in progress)

Carol LeNestour - Ontario Hydro. BAND

The challenge facing many of us today in systems engineering groups appears to be how to accomplish morewith less. In our industry. daily, we face increasing ~ds., both regulatory and imema1; increasing economicpressures resulting in effons to maximise system efficiency. reliability. and safety. on going consolidations; downsizinll.and increased individual resporwbilities. -

Cne of1he key roles ofa plant syst":ffi engineer is to ensure that their systems c'lntribute to the overall highreliability oft:Ie plant. Engineering system survemance, the tracking, trending, walkdo'WDS and genenl systemmonitooing perfOllfled by Ihe system engineers. provides a foundation on which to build an effective system engineeringprocram. But currently Ettleind~ guidance exists to aid system engineers in determining what is an appropriate leve!()f system monitoring in order~isc the system performance obtained for the CIlgincering~5 in':eSlcd

Faced with these issues inNS station, Bob Waselus ofSouth Caroli!la Electric and Gas, V.c. Sunmer,submitted this topic as a candidate task to Plant Support Engineering at £PRl during their annual meeting in-June 1995.As their mission statement states.: "PSE b. a u.tility driven support resource.. wh~ objective is !o suppan utilities inreducing O&M costs related to engineering while improving or maintaining tedmical '4ua1ity." Based on this mandate.the PSE subcommittee approved this wk. and a System Monitoring by System Engineers task group was funned.Mcmbus ofthis task group included Bob Waselus - our chairman. Leonard Loflin - the £PRl program managa', the taskcoot:ractor - Duke Engineering Services, and industry and utility representatives ( system engineers. superviscrs andmanagers, including a representative from INPO).

The challenge that we were handed at that first meeting in Fd>ruory. our charter. was to, by the end ofthe year:

"Produce guidance uSLful to "Kliv;dlloJ system mgineers and system engineering organiSQtions inOlX.:mrplish""mt oftheir responsihility to mon.tUT system and component perjonnanCt! to achieveapproprialt system perjonnoJlce.

The Task Group is to search tire industryfor best practices und Je:;,wtJs l"UlT1Jed that would be of:mT'Jed;aU benefit to system engineers. Particular emphasis is to be given kq parameters andindicaJors. proloen proct5Se.... techniques olld technologies tloat are specifically ejJectft-e in obmil1hrggpprt?nriqlt ~srem pe.1onnOlUe, whit'! mi.',i",ising the COIISrI.7rptio., ofengi.'1f!ering reSources. ..

In silon _ to optimise a system -.urveilll'nce ptogram by balancing th~ ensineering effon expe:1deJ and the vahleof the rwtant performance improVemt'11t.

Ol.:r first task was to c'etennine the "Stat~ of the Union~ 500 to spe<ak. We did :his by developing a. survey whichEPRI·PSE sent out tv 87 member utilities_ Wi,h this su.vey, we attemF-tcd to find out not only what the utility wasc;urrently doing for systeM r"onitorn:g. including any best Fra=ti-:es that they would like to share. but also. what form ofguidance that they would like to see.

In the first meeting, some of:.LS had enviiioned tha: a large pan of the guideline could be obtained from stationswho were doin6 surveillance well. and t"-t through discussions. site visits and survey results, we would find what the-:urrent irKlumy best practices Wel,e. We discovered that:

• Most stations are perfonning some levei ofsystem monitoring. However. most stations indicated that theirprograms needed improvement.

• There a..-e:na:JY incorn::i!terh-i.es within a plant. within a multi·~itc utility. and within the indusr::-y:n general. Noide':!tical programs were ~nco",erM. Some S[ations may be monitoring 100%, oftheir sys1":~ others may not bemonitoring any system!'.

L ---------.J

(


•

• [t was difficult to wrrelate the scope of the system monitoring program and the overall pert'onnance ofthe plant.For example. those statiQnS with good SALP ruings., low O&M costS. or high capacity factors do not always havethe most e:aensive system monitoring programs.

Based on input received from the survey. along with 000 and SALP ratings in the engineering categOries. cwoplants (Byron and Limerick) were chosen for oo-.site interviews. These on-site visits. combined witb results from thesurvey led us to the conc:Jusion that we needed to develop our own process for system. monitoring, a process that wouldintegrate me comments from the surveys. the rMuhs of the !rite visits, and the lcoowtedge and expelace of the taskgroup~ to provide guidarJ::e that could be used to develop, improve or validate system surveillance programs.

What we have arrivtd at to date, is an 8 step process focusing on the critical dement: ofan effective program..1D order to provide some validarion fOT the process. it was tested on five diffiR:rn systems at four r.uclcar stations. Thefiatt t:1at tine of the~ were rompleted by system engineers who were DOt involved with the task. group. helped toprovide. gnss roots. <01d bolly =icw. .

.5JUtm Mppihtripr Prprram DcYJIopmnt Prprm

SUp 1-........ Scope Der.._SiDce nof: an stIIioa systems wiD rcquirt monitorinJ; at !be systems level. this step is jusI: ad iDitial dWsion as to

wbothor ac_.system will be _ in the program. the...,.J Ievdor~ will be _ ....... 10

mUiag tbiI decisioaeenaiD c::riteria sucb cost. pl'Oductioagoals. rctiIbility recp.airemeftts iDdustry cxpcr" DC:' and as~.......- ( ..~ NRC', _ Rule 10 01l50.65 "Requiraocors fOr moniloriq lI>e J:ffiocUven<ssofM'd" It NucIeor P..... Plants") must be coosidaal.

Step 2· »dille PrrformaDct GoabI In;1ica1011 for Sys1tmIn~ step. we ue iooIciDg for parameters tha' will tDe1LSL<rc the dI'ectiveness ofthe system monilorinJ: ,rog:am

-. key daa1Cllt in ensuring.continuous UnpiOJt:n:len:. These indkators cookI be di:rcct incfu:a!ors such as S)'StCDl

availQlity. or io1di:cct intficator; such as maintenance cost for tbe S}-"Stem. It is essential to note here t.'Ie.rdariooshipbeta_ m. _ soaJs. and the ovuall planl soaJs ( sud! a:; pP>duai.., cost "'" saI'c:y_I, ..... 1<no'M'8 theimpact (Ifthe system on the plant will help to achieve the appropriate IcveI ofsyftm monitoring.

Step 3 - DerIDe lmportmtce ofSysttm FIIDcti:ons and CompoDcctsThis is the- aitical step in foc;o.JSinS monitOl";ng effons ~ t~ system functions.. whore modes and 5illure effects are

defined. as well as critical i:tterfaces too~ syStems. This determination offaihae n.odes and effects is used top.ioritise the monitoring effon. TOis infonnati.Jn is avWable from such sources L" d:sigD basis documents. re'Jabi!irycentred maintenance assessmer-ts, and probitbilistic asstSSmeIIts ti wea as pllUH e:<pefletl\:e.

Sttp 4 - DerIDe Systan Manoriag Reqairemeats:n order to de".enninc: "how" and ""what.. l.O DKlmtor. degradat:ioJc mcdwusms and ind:C&Lon; for the failure

modes ofcriti.:al system fi.mctioos are determined. foc eumple, llt ttte componentleveJ. ade~nmecllanism for aheal exchanger- is fouling.c!~n iudieuon could ir.clucle increased pres:sure crop and a deauscd temperaturechanse. Here is where the generalist n8n:re ofa system~ CQrnCS into plaj'. since tNs step requires a thorough~ of system func:tions:. system equipment, induding its physical. mechanical and electrical propenics; as well asboth short term and long term .gang and wear proc:esses. The guidellne will providll: support in these areas bycontaining sample survcillanc.: pian shells for approximalely 5 seoteric systems, as well a::: n:ferences to other :;ources "rinformation.

Step 5 • Idutify die Ibta RequirementsIn this step. the :o,'Slem engineer ddinc$ the data type. acqt.isition frequency• ..net precision reqcired 10 monitor

degradation mecbanisres. ANoJysis ofsystem performance may rCC\uire integration ofseveral ditrerem corupcnent treI'ldsto determine the cumulati""e effect on a S)"Stem. Fa!" example. you may have a systemwh~ all the individual paramet:rsarc opcntin£ in • degraded star.=., but within the ~lcnble level... but the systCl'tl as a whok Qa)' have unacceptablepeaf'l>i ....ce. (e.g. a valve slow to stroke. combined with low tlow rro.na~. and beat exdwlger fouling)

22

The majority of the data required to monitor system performance is being; collected at most nuclear stations byvarious different departments. the key is to interface efficiently with these depanments. and effectively integrate the dataAnother key point, and a trap 11m many system eng!neers fall into. is to avoid reviewing: and trending data that does notsupport the monitoring for degradation ofcritical system functions.

Step 6 - Identify Actions RrquiredSetting appropriate action levels. and understanding and documenting the action to be taken when these ate

exceeded should allow proactive intervention to prevent failure. There were many examples in the industry whe:e datawas being trended, but acceptable limits, either absolute vaIues or rate ofchange. had not been established. and actionplans did DOt~ t.ction plans may contain such activities as increased trending. monitoring or testing; root causeaoaJysis; d~gn revi-=ws; routine maintenance; or operational adjustments.

Step 7 - Estat..lish Communication MdhotbAlthough it is clear that the system engineer JIIl:st de6ne all conmunication chame!s in order to keep the

required technical information :!lowing. management reporting of surveillanc:e results is also enential in order to ensureL'W: system problems rtt:eive the appropriate: level ofatteJnion. One ofthe best practices that we found in the industry.was the usc ofa"'system repon carer. or a "system health sheet", Most syscems are assessed in several aretS, includingp.m"'..... (roliability and avaiIabiIityt, dtratings; .....enance baddog; physieal oondilloD: operator w:>rk-aroundS; anddesign issues. and a "'window" or lJUlUllciator colour is usuaDy assigned. These repon cards arc used as a tool forfocusins plant resources, since tivcry issue contnDuting to • window alann teqUires an action plan.

Step I SystecI Monitoriag DocameotatioaIt "essential that !be system engineet dO<umenl the decisions made ill !be develo......' of!be prognm. This

dOl.:Ul'dellt. which should become a living document, will provide a eurrer:t and a historical technical basis for theprotram. and an invaluable tool when transferring system r.:sponsibility to another engineer.

Future Plal1S

The final meeting of the task group is in December. At this point in timt. !he report should be 5naliscc!. and~ with by the entire group. Copies should be a\lailable in FebnWy or:xt year. Planr $oppcn Engineering~obtained funding to produ:e approximately 40~ "shells" ofsystt:m surveillance plans - shnilar to :he few providedin the guic.leline. AJthuu~ these will focus on PWR and BWR technology, there may be some valuable infor.nation inthese plans that can be applied to CANOU. Work will start on these in January.

FPRI is also planning workshops in Mayar.d August ofnext year- to "rell- mIt"' the guideline., to providesome additional baelup-ound and insig.'lt into the "Haws and~- oftne process, and to discuss operating expi:riencewj:h the progr.un to date. Currently there~ three uti1itie~ implemMltinS the draft process.. tnd th.eir feedback, alongwith ethers. wJ1l be disccssed at these workshops.

The nc:id test ofme program \l,.il1 be proven over rime. however 1 think that it is very encouraging to hear thecommentS ofsome ofthe system engineers wt.o were involved in tbe original b ials of the process. "With about the sameeffort.. I can now monitor tlWtgs tlw are more consequentiar. 3lld" Now I have a better understanding ofwhy I'm doingwhat I'm doing" - all keys to a successful system monitoring program.

System Surveillance

(

itqJ

5_AI-

5,... B

Program5_C

Scope DofiDilion 5""_2

5_0 Doli... Performance~

Defme Import2llceGoals 1!DcUcdors ofSystem FundiollS I-

for System and Com_

-c... ·Re1iabiticy-¥roduaiOD &oa1s -o&M....ou ·RCM.IDdusUy experieDc:e ·UaavaiJ,bility ·PRA...._- -:teeulatcnY inputs -DBDs

-Plam: experience

_.SlJF$

_.t.. Der_ Monitorinc IdoadfY Data IdentifyA_ '-~ ~ Eeqv;red

i._- .s-.. -Inc:rc:ut trelldingc- ._- Fie L ) -/.djust pis I indicators

-Re:dutdoa metblXi -AdjBst lests I m''n'....'Dee-TP.lDdinl -Design review-~bUlds

·ActioIIlevels

Slq>1_.

-I . Esttblisb System M~nitoring Perform Systemt.-. Communi<ation .. _-~Melhods

Documentation Monitori:lg.-...... -Basis for monitoring-om ............-System rqxlII. eards .-In(onnaI comtrI1miaooa

AGURE 3-1. SYSTEM MONITORING PRCGRAMDEVELOPMENT PROCESS OVERVIEW

-'--

EXAMPLES OF VARIOUSCASES

•••••••ll!Im

•24

• A case study on the implementation of anintegrated maintenance managementprogram

• Darlington equipment aging management

• Pickering thermal performance : Low techVS High tech surveillance - practices.

II Reactor noise analysis application inOntario Hydro.- A statistical technique used for surveillance -

System Surveillance

A..·.·...\i} EXAMPLES OF ONE CASE

The Darlington experience showed regular weekly and monthly routines oncritical areas is essential. Sources ofproblems found to be:

• Leaking blowdown valves• Leaking CSDV's• Reheater drains flow• Condenser performance• Calibration of instruments for reactor powers

•

Darlington Equipment APing Manarnnent

Nuclear Plant Life Assurance iNPLM

Abstnlct;

The progIlllll has two major thrusts:

a) Monitoring degradation ofcxpensivelnot easily replaceable pieces of equipment suchas Pressure tebeslboiler, etc., and,

b) Preventative maintenanre ofcritical pieces of equipment (replaceable) such asvalves, pumps and so on.

The first group is well underway for routille inspection every four years. The differencebetween Darlington and previoRs OH stations is that IJaseIine5 are done within five yearsof operation with an emphasis on deteeting small changes so that a rate will be deteI1DinedbY year 10. This means going beyond regukUory requirements and doing wpecUCD.< withmore scope and mol'> precise tooling. For example, several boikr tebes arc removed todetect degradation < 5% through wall. This a~cUI"K1' is not possihle with eddy currentinspections. A comprehensive program in this NPLA area is jUdged to cout':'ibute 10%reduction in incapability in later plant years.

The second group of equipment involves abo"t 2,000 items. ;,ach one, if failed, eitherca=s a loss of production or requires II unit shutdown to repair ()r rel'lace. Systemengineers have idenlified these items and callups are being put in place to iJlsl'ect andoverhaul as required. The typical tilDe fr-..me for these are 4 - 8 year intervals. Severn!tools have been purchased to provide effective m:wuenance such as valve monilOringdevices <FLOWSCAN/MOVATS), thermography, elaetomer tester tool. vibrationmonitoring, HX tobe cleaner, geneIator/tu.-bine inspection tools requiring no disassembly.We are also considering a portable skid for nuclear HX shell side chemical cleaning. Webelieve this equipmentmainte~ce will achieve 5% reduction in incapability.

l-- ~

26---------------

System Surveillance

Darlington Equipment Aging Manamrumt

Introdoct!on

As a new station, we had to take stock of what OlIN stations have done in the past anddecide if those approaches would work for us. A review was done of the causes ofincapability at those sites and FJgUre 1 was our conclusion. We believe an annual capacityfactor above 80% after 20 years may not be achievable. In the early days of Pickering A,we postl'.d 90% capacity factor but after a dozen years, 80% was demanding. Whenpressure tube-s came to the forefront. capacity factors plummeted. Many of the olderstations are suffering from = genera10r problems. PointLeprean, which had anadmirable record of production achievement. appears to be experiencins some snrprisesafter 15 years.

We decided here wilen the first unit went into service to aim for an annual 80% target anddo the required inspections and predictive mainJenance to get an early trend of equipmentperformance.

1.0 ~

We have Ihree units in Engineering Services thaI devote tbemselves to equlpmenlissues. The first ,.,nt is Nuclear Plant Life Assurance (NPLA) and focuses on theexpensive, hard 10 replace items. The o!hcr 2 units devote themselves to I&C, aIldmeehanjna1 equipment.

Wc beve programs in place for periodic in.."J>eCIions of Steam Generators andPressure Tr.bes. Every year we inspect one unit with an accent on more mbes orchannels ai1d utilization oftecbniques tba! fini! SIIllLI1er defP.Ct<. We use both tIT andtube removals on SO's to cb:1racIerize early 3igns of degradation. For pressurelUbes, we have 3 devices 1hat are FneI1ing M'lClrine delivered; a camera for liner/e"dfitling internal inspeaion, an ultrasonic tool ior pressure lUbe defect detection(PIPE), and " laser detector (OPIO. These last 2 devices enable bettercbaracteriution of flaws at the inJp.ts.

Plt.ns for next year are Caland'ia internal inspeclioo and cable monitoring. We seeCal3ndria problems as the next majol i.'lSue facing the older CANDU units and wantto gel an early start on a program.

•28

Our piping progmms are deficient in that we do not have piping surveillance forcorrosion nor do we have fatigue monitoring for high energy piping system. A"new" approach is being pmsued with Ontario Hydro Technologies to developmechanical fatigue probes for use next year. These devices existed 40 years ago buthave been obsolete for 10 years with really no replacement. We have secured themanufacturing capability from a defimct company and will make our IIfSl batch nextyear.

2.0 Significant Equipment Mainten;mce

This is a far more difficult subject. :1 is never clear where to draw the line becauseone gets the sense of we can "recover" if equipment is falling apart in laIl:r years.The NPLA issues have -buy in", but on increase in mainlenance either by 0verbal:Js,orp~mpted replacement is generally resisted because of~etedresourtes. Firefighting takes a higher priority over long term issues. OM&A l'.Ild staff numbers areIiJred;" in fact there is an expectation that it must trend downward to be economicallycompetitive in a chanting marketpIace.

Before mainrenance staff are consumed wilb corrective maintenance, we conscio:JS1yfilled their plate wilb a high preventive maintenance workload. The followingprograms are in place \ly the Production Support Unit, and lbe other two u:tits inEngineering:

I. Heat ET.changers and Steam Generators.

- nuclear HX inspections, and a possible chemical c1eaJrlng ofa SID coolingEX.

- cleaning ofconventional IlX's a minimum of every 4 years.SO water lancing every 4 ye&rS and chemical cleaning every 10 yem.

2. Major Pump Motor Sets.

- disassemblyfmspection of a sample ofPlllIl!J8 and motors begirutiug at year12

3. AOVIMOV program doing routine MOVATS and flowscanner Ie.tlng (specialistfunction) OI! critical valve:>

4. Routine oveI:la:J1 of critical switchgear.

5. Routine thermography of me<:\Janical and electrical equipment (specia!istfunction).

6. Vibration Monitoring of 1200 pieces of ro!ating equipment (specialist fuaclion).

System Surveillance_.__._--

7. ElastoDler Testing.

Currently this is an Engineer part time. In the longer term,. we see it as amaintenance specialist tool

The above techniques are applied to equipment, that in the Engineering-Maintenancecommunity are commonly considered to be cost effective; ie. not necessarily on criticalequipment, but on equipment judged to fail SOOI! ifa tittle effort is not put in to diagnosisand repair/adjust.

2.1 Critical E«Jpjgment

The following describes the approach:

1. 170 systemS evaluated for risk of Unit/Stanon incapability. Expert judgment usedcombined with bi.story of similar planm. About 35 sySlemS bigbligbted.

2. Sys"'..em EngiueerslCoordinator evaluate key compcnents in each system hy w:ingtlowsbeers. Single failure resulting in downtime results in designation as "highlycritical". Single failure resulting in si~ficantloss of redundancy was considered"medium critical".

3. Control MaintenanceJMecbanic3l Maintenance pei'SOnnel eva!nate =riti.cal list oneach system anlj deteImine lik.eJy modes Clf failure from experience <'1 other sites onsimilar equipment Preventive y..aintenance tasks recommended

4. Call-ups put ill place. spare parts ordered. support documentation inie:i2.ted.

5. On-~oing surveill3nce review by Engineers of success ofprogram.

NOTE: 1) In addition to above. a parallel review is done across all sy.;tems on keycomponents - Heat Excbangeu. AOV·s. MOV·s. Pump-Motor sets. Mostimportant equip:nent given "prerlietive.. maU.tcn:lDce.

2) A l'eJ>ara1e proc:ess is used for station "life threatening''' equipment. SteamGene:.:ators. Cabling. Pressure Tubes. Piping. Major Civil St:ruebJreS. TO,etc. (> 50 M$ or > ~ months to repair). Each is assessed for de,grr.datior.mechanisms and a periodic i&-pe~on plan pot in place.

The review dest;ribed above bas identified abo"Jt 2.000 pieces ofequipment for \lTmchcallups ate being put in place. The normal time frame is about 4 - 8 years for firstinspection. This program is scheduled fOl completion in 1997 after which the monitoringfor succes&'failure begiDs and adjostments made on an ongoing basis.

•30

2.2 Balance ofPlant Equipment

To preveot unreliable operation in the 25-40 year range, we need to replace classesofequipment which is either difficult to maintain (fails often), or costly to maintain(long repair times), or simply cannot be IlXed because of no parts (obsolescence).

Appendix A provides a best guess of the materials and staffneeded to do it. Wepresently are understaffed to do bulk Changeouts and cannot see that Ibis will everchange. However, what can change is anIn= in efficieacy in maintenance.Wrench time is low here as in other OHN statlons for many reasons and outside thescope of Ibis evaluation. Because it is so low, it presents a realistic opponunity.What the reasoning shows in Appendix A, is that a l!oubling or bipling of staff toeffect such refurbisbment is uneconomic. We either accept a lower capacity factor(3-5% realistic, 10% upper bound) or we must increase maintenance productivity.

Ifwe go back to Figure I, we need to identify what eqnipment requires maintenance.This has been arrived at in two ways:

a) Critical equipment evaluation which was described in 2.1, and,

b) A tabulation of the most 1l1lmerous equipment types in the plaitt. Tuis is shownin Figure 2.

Frem Ibis follows assigning engineering resou= to monitor these types. It is :lotsurpril;ing that we are targeting hiring ofat least one person with specialty in each ofthese areas. We are alsv <!eveloping staff to cover the£ areas. These are Dotteday's problems but there is a good chance that they will be in future. (ToC"cmplete the picture of what today's problems are, Figure 3 and Figui-e 4 areprovided).

Summary

The success of an NPLA program is measured by the lack of major e!}uipmem .-mpri.<:es.Not surpr.singly it is, therefore, a managed maintenance program a:td not some ohscureback office c.tercise. It is field work on the right things, the right amount, and occurs atthe rigbt time

Darlington is putting in the eft"ort to try 10 get their maintenance plans in place early in lifebefore areactive mode sets in due to smprises. Figure 5 is a summary of where we are.today. It is incomplete in that the discussion above highlights other component classesthat should be added. (OUr assessment is that it will take a few more years to complete).It also encompasses other issues deemed important by 3.'1 OlIN tearn reviewing Ibissubject (eg. system surveillance).

In the end we will get there with perseverance and a bit of luck.

System Surveillance

(

(J.:- I'..- .. ~.:'

Figure 1

OPPORTUNITIES

15% LOSSwmCHISAVOIDABLE

100%

65%

75%

NORMALMAINTENANCE

DOWNTIME

MAINTENANCESTRATEGIES

NPLAISSUES

OPERATINGCAPACITYFACTOR

(EXPECTEDBY YEAR 20)

I

-VALVES, PUMPS,MamRS. HX'S

PT's. SO's, CABI.J:SCALANDRIA. PJPING.CONTAINMENT

L~

The ef'fEd. of a 0.5% improvement has the foIowillgpotenlial impact

TPl- is C91aJ1ated for each unit, as well as for theS1ation-a_.

THERMAL PERFORMANCESURVEILLANCE AT PICKERING ND

.....•-..

Measure:TargetCurrent

CYCLE SOURCEGoat MaximiZe por.w:r transferred

to secondary side.Thennal Power Error (TPE)-0.5% ~ TPE ~ +0.5%TPE= +0.03 %

a major aspect of secondary-side operation, and eachof which has its own measure and target• the cycle soun:;e TPE• the cycle sink CVE• the turbine cycle (ie: everything between) TPI

DiS'cussionCtose control of TPE means close control of theturbine cycle heat source. It essentially invCIIvesensuring Ihat we are delivering from t.'le HTS 10 theseoonda:y coolant every Megawatt of powerwhich weare - by rlCe:1S8·~ to transfer. This inYOlvesroutkle calculation of each unit's calorimetric. andoccasional adjuslmenls lD the Digital CcnlrolC<lrnputars' Reactor RegulatIng Syslem suIlpn>gmn,~ order lD-., c:Iose agreement belweenindfcaled and actual reader thermal powers.

Current StatusWe have optimized this by means C'f a rigorousprogram of testing, OJA. fine-tuning of FPTRs, andmi,"",izing unnecessary ,,"mary coolant loads. TPE isafforded highest priority, because 01 its potential safetyimpact. In the laSt 5 years we have satisfied thetargets. TPEs in 1994 and 1995 have l::leen near-zero.However, the ament process is paper..mtensivt= 3nd1abour-:nienSive. :nvoMng daIa ooIection lr/0perallnirom complf.ers, C-ontrot Room. and fieldinstrull'lef'tmion, folkJwed by off ·1i.1e analysis byTherm~namics staff. This paper process is asstreamlined cr:; can be reasonably achieved.

""I1

1.5°t,,..;xlt

1... 1·················································

I D.03 L....1=1 ..i.,,+: I-uu_uuuuuu

,,, 10m ., 3m _ 5130 _ 70 ll/2I!lI27' 10z:71tGl112C!li i(2.72 tifWe p~r u"e)

x (a7SO hourslyear)23.85 GWhfunitlyear

Thennal Powe< Em>r [% JCondenser Vacuum Efficiency [ % ]Thermal Pt'rformance Indicator [%]

y..,. TP'-1990 2.59%.1994 2.01%.1995 'i.75%

0.5% =:>

Summing the deficit of each cJ these from its nominalvalue (0%, 100%, and 100%. f<SIlOdively), ""'_Iquantifies how far trNaY we are from the entiresecondary side's design-level of operation. _g0.5% latilude (and C>:IUIlting negative TPEs ..zero) oneach measure, we are targeting performance within a1.5% deIiciI trom desig..-level 01 operation. Thiscumulalive deIiciI trom des;grHevel operation lin beennict'.named 1he -rPt.-- by Gel terating Units staff, andhas been incorpotaIed into the Genelating UnitsManagers'. perfonnance contracts.

THERMAL PERFORMANCEThermal performance meilSUrement at foND is brokendowr, ;mo three components. each cI which addresses

SUMMARYPND has three leadin~ indicators of thermalperfotmance, er.ch independent of the other. Theseare:

TPEeveTPI

IN.M. Cichowlas, P.Eng.Thennal Performance Engineer

Engineering Sciences UnitPickering Nuclear Division

The 0.84% improvement of 1995 over 1990 is worth4.6 ~lWelunil. Assunring a station capacity Factor 0180%. tt'.is works out to approximately; 257 GWh/a. orabout 5.1 M$/a.

At Piclterilg NO. a decicated program of thermalperfDrrnance improvement has been in ptace smce1989. Available performance data ooes back &t least 6years, eg:

•32 System Smveillance

unit~ differences in oper.tIing intervals andpcwer IeYeIs could bias the results. Explanation is bestdone by way of 8/1 example.

In the following tabJe are ~luslr.lted - from left to light the following infon'llatiCln.

Unit [0] The PHD Unit numberTP'- !%} "Thermal Perlormance Index"

(Itlis is the deficit from design-lewlofoperation. and is calculatedcirectIy from TPE. eve, TPI.

a.TPIc. l%l The improvement in TP'- MNn1~m 1995

NET {GWh] Neteledricaloutputforeactr unit,1n199S

. SOW IGllllh] The enetg)' saving in 1995. basedon 1he difference in TPl..betw6en 1994 and 1995. This Israk:ulated by:SAV" NET x (ATPIdex)/100

1.. 1!l9:5Unit no-. '""- ~""'- Io!ET SAV

1 2.77 '.&7 .o.tO 1982.407 -1.982

2 Us rW nIa - nIlI) 1.47 2.08 .o.~9 21i8~."';1 -15.82.. 1.92 2.05 .0.13 Z773.esr -~.606

~ 2.,~ 1.02 +1.11 3357.1Ill6 +37Z1, 1.$7 OG +ll.711 .3483.!i56 +21.17

7 1.18 1.70 +0.16 404$.0&5 +6.479

@ 8 2.$4 2.46 +0.08 4Oll3..ccw +3."",._.~- .

Avg 2.01 1.75 +0.2£To:at 22331-726 ~.n

CONCLUSIQNS• No mad« how .::oRt-effeclive initiatives in themlaI

pet'f0l1ll811ce improvemel'lt are mown b be. theyare stil discietiollary. When resoun::es are limitedand Production priorities irwotve minirr.izing outagetime. continued ope~n, and iS3fety-l'efateissues, lhermaI perfoml8flCe inltialNes will beshelved.

• consequently. engineered c:hanges which involYesignilicant bloclu: of time<ommi!rnent by variousstaIlon worit glDUpS Will ~'8 over extendedperiods rJf tim£.

• In a tmJIIi-unlt em'imn~nt, pmjedS involvingsignilicl;nt engineering~ can aD !Je irlciffela\t 5tlgas of implementation. TICICkingprogress and ptOIIiding support is not so muchdiflic:ult, as it is awkward and time-consurning.

..~ Changes are often installed bYuniH'espOnsible crews. ThiS means that sepa.-ate

e-..

Work PlensiPackages. materiel management, PAljob Ofientation, and support must be provided for~1JlIit.

• As such engineered changes can strW;h overyears. peiSOIlneJ changes mean that job brieIingsmay have ttl be c:ondllcted seYeraJ times.

• The ~svltof the isSUes descritJec:I aboYe is !hatsignificant engineered changes can take a greatdeal -:A time to show any return on investmentDuring the time of installation, the process is aSignificant dr.Iin on engineering resources.

• The Iow-tech approach s1ands a much betterChanCe of bP.ing completed quicIdy aIld~.

• Since it gets oft the ground ve!Y quicldy, a Iow-tedl.soIutitIn begins tD ltbow resulls amt reduceprodUClMty los.5es wl\ie more amblllous programsawait JorIg-fIlml irrlpIernetltr..

• The PHD "cycle isollllionlAlam trap' projed wasfully ecnfigrnd by UftlHPeciIic crews within days,ar.d was done an • pick-up b8Sis. The unde~analysis W3S dane by a central seM::e>JrgllfIizlition, and so was n:lt affed8cI by stlIIionresource c:onstrai.'ltS. 0YeraII sec:ond3ry-sidesurveillance is lOmiIally being il'lSlibJted 00 a KlSSbaSis.

• By keeping SUfVeiIIance requirements quick andsimple. !hesu=ss rate kif~ routine elrealti:Jnof the caIIups fcf the surveillance by Ge:walirlgUrms staff is high.

• Analysis of cycle isoIationIsleam trap surveillal'lCeis perfomJed by I:/'te~-namics group. Bycompleting the &natisis quickly and feeding it baCk10 GU staff fnr prioritization of repAir. lhey hawr.lOfe or~ instant gratilication. as well as thecorrect~ tnat we are provlding them asel'\rice. This further PCIds h !heir level ofin\-'oIwmentand satiSfa.::tion.

-------------------------

CANDU Systems & Equipment. Surveillance Programs COG Workshop November 1996

34

REACTOR NOISE ANALYSIS APPLICATIONS IN ONTARIO HYDRO,A STATISTICAL TECHNIQUE USED FOR SYSTEMS SURVEILLANCE

D. GlOckler D. Cooke, G. Czuppon, K. Kapoor

IteactoJ: Safety and Operational Analysis Dq»attmentNuclear T~ology Services, Ontario Hydro Nuclear

700 University Avenue, Hll-E26. Toronto. Onta:rio MSG lX.$

M. 'I'uleU, D. Williams

Electrical bstrwnel:lt aDd Control Systems, Pick:erir:lg Nuclear DivisionOntario aydro Nuclear, Pi~. Ontario LIV 2R5

ABSTRACT

Reactor noise analysis is a. non-iDlrusive statistical tecl:nique re:gola.!-ly used in surveillaD:e andUiagnoeti.cs tasks. The pa.per CODCf!:D.uates 011 some of the receJlt applic:atioDs of reactor noise auJysisin Ontuio Hydzo', CANDU lIlatiaas, related to the dyn&lllico of iJ>.co>e lux cleie<.... (ICFDoI ...dion chamben. These. applic.atioDs include (1) detectiD.c anomalies in the dyDamics of ICFDs aad ionchamben, (2) e:stima.tini the dect.ive prompt fractiOlaS of ICFDs in power rundcnrn tests and in noisemeasurements, (3) deter.tiDi the mechanieal vibration aI ICFD instl'ume:bt tubes irLduced by moderatorlow, (4) detectiJ>g the Dle<hanical vib..u.. of fuel <IwmeIs ihdu<ed by coolant low, (5) ;deD~the callCe of cz.teSSive~ lIuciu.uloDs in =tain lu de"..e-.:wra, (6) valiUting the dyPUli.; coupl;q:betwee:D. liquid ¥one COIlttol.signals. Some of tbase applieatioDS are performed on Q regular basis. Thenoise analysis progu.m, in db: Pickering-B station alone, has saved Ontario Hydro millioos of dollarsdu:tiD.( it,s first tbt« years. The lt$IIlts of the 1lC»5e an.alysis Jl'lOgam bal'C heeD also reviewed by theAECB with fav«a.ble results. The AECB have npressed interest in Ontario Hydro further exploiting:the use of ;:I.cW;e analysis technology.

INTRODUCTION

Reactor noise analysis is &statistiw t.ec:hniqo,Je for extracting iD{ormdiCG OD reactor system d)l12:mksfrom the :8uetuatioDs of instrumentation si:nak m~'11red dui.ng study-state opera-bon. The SIllaUand rneasnrabl~ fiuct.u3ti:ons of process rig!lals a.tt the results of st.o:hast.ic eaed$ inhere4lt.m pbysic.alprocesses, web as heat traDsier. boiling, eoo1mt flow t.U1'bulence, 1issioJ:l "rcc~. structwal vibrationsand pNSSUl'e :'>$CjJlations. The~ of reader~ analysis is to moni~r and assess the conditions of~hno!ot;ieal J'rocnses anel. theirins~ODin the nur-!eal r~A)I in ::l,. nOll·ia.trusive pauive way_The noise meaauremc:a\S w: -osually perfOIniCd at ~tudy-state operaticm.. wAile the availability of thesignals in their :tespeeted systems (ei. shutdowD lo-y5tems,re~ system) is not intenupted.. Althoughrea.ctol" noise analysis teebr.iql!e5 usually offQ an indirect way ofdiagnO£tics and zequire expen.l::nowled&e,often they are the only diagnostic indicators of processes ina.ceessible to direct pha.t testin&.

In J992 &l\ ext.eusiv't program of reac~ Doi.~ analysis W1lS initiated in Ont&..-io lIydlo to devtlopJIOise..based statt¢iea.l teWn.iqaes for monitori:tg process aDd.~t.Won dynamics, diag:aostics andearly fault. dek:c.tion. Since thcD, vanODB CANDU-specific noise analysis applica.ti.ons bave been <level~ and valida.ted. The ~bued statistical teclmique5 are beinr; sucocssfu1ly applied as poweriult.MUbleshooting and diagnostic: tools to a. wide variety ofactual.ratiorlallkC problems. The dynamice.hara.et.eristics cl'~ plant. ecmponents, iDstrumalt.aticm anel pr<:cesses an UlOllitored 0Ill'\ legular basis. A CGmrrehensive "uoise snrvey1' cl' d~t.eeior sitnahl from the stat..dard. ir.st'rU!Jlellt~D otPicke:riD&-B,Bruee--B and Darlingt.cm. units ha'o't' been carried C'ut in the past fow: yu.rs at various o~"iliLg COtldi

tiODS. Also, :oeccIIUXlCDded rtandalds and pro-:eJures for regular station noise measurements have beendevdoped. II. these measurements the fe.lsibility of ap,lying bf;lise analysis techniques to ad.':lal opera.ting data bas been clearly demonstrated. The results u..ccat.ed that tM: detecuOh and characterization

System Surveillance

(

OTHER APPUCATIONS OF NOISE ANALYSIS

Noise analysis has been successfully used in pres&1lIe and flow measurements of~ primary "beatuanspon (PBT) system too. The a.pplication includes the following~ (1) estimating the responsetime of pressure and ftow transmittas and validating their dynamics, (2) identifying the resonance frequencies of .""""'" .....me !iDeo, (3) w1i~ FINCH flow aAd SOSl safety low oigDaIs, aDd (4)characterWng anomalies in 8.0_, such as signal dips and osc:iUations [13.I4}. Noise analysis also providesa non·intrusive metbod for moui~ring and estimating the dynamic response of KIDs installed in theprClCes!', and fOl' isolatin.« the c&\\5l!: ofRTD signals auoma.Iies (spikes induced. hy ground fault detectors).BoiliDg in FINCH fuel~ can be also deteeted by Doise analysis. The detection of coolant boiliDgin FINCH fuel chllDllels is based on the measurm1ent of w« anel outlet ftow BuctuatioDS. Noise measurements in DarlingtoD "showed strong conclatioD. between the oteunenee of boiling (iQdicated by fuelchannel otItl~ temperature) aJ:lQ the eo.bexe.nce and phase fm:u:tiObS of inlet and outlet .flow :fluctuationsiII the hqueucy range of 0-1 Hz [3}_

CONC1USJO~

CANDU nou.e measurements carried out in the past four years provel.1 that tault detection aDdvalidation of proeessJiDstnlmeD.tatioD dynamics can be based on the existeD~ of multi-dw1D.el complexpatwms of statisti..:al D"'Ue iLgDa.ture$. The teclmique is be.iq suaessftilly applieclllOW' in a wj~. W&l'id:yof 1la:.1al station problems .. a powetful Uoub1ef.hoo~and cfia&nostit tool.

•36

ASSESSMENT OF FINDINGS•••••••llm

• From the examples given and the topicscovered to this point it is obvious thatmanaging the various requirements is amajor managerial challenge.

• For running plants lnuch of the data andtrends required are not readily available.Changes are expensive and discretionary.

System Surveillance~~ --- -~- ----

(

ASSESSMENT OF FINDINGS•••••••IID~

• In addition to the types of infonnation thatcan be taken from an instrument or alocation there is a vast amount of 'corporatememory and experience' that is spread overmany disciplines and people.

• 'Maintenance records' - ifwell recorded canprovide vital insight into system andequipment performance.

•38


• The majority of the instrument monitoringis indicated in the main control :'::>om but itis mainly instantaneous and only availableto the control room operators.

II Maintenance staff 'give up' feeding backinformation because 'nothing ever getsdone' - sub performance is accepted.

System Surveillance

(


• Historically system surveillance has beensubjected to severe budget restraints,however t.'lere is evidence of change as thebenefits become more evident.

•40

LINK TO FUTUREPERFORMANCE

•••••••IIO~

• Considerable effort is currently underway toapply technology to system surveillance tomake it more effective and improveproductivity.

II Various elements ego vibration monitoringare quite advanced and the integration withother monitoring such as chemistry isstarting to show promising results.

II There is no 'magic bullet' for surveillance.

System Surveillance

(

() LINK TO FUTUREPERFORMANCE

_ ••••••IIIO~

• The intellect and experience of the staff iswhere the greatest 'value added' isobtained.

•41

F'......"'.'\<", :}

•42

EFFECTS OF LACK OFSURVEILLANCE

•••••••llm

• Dropping capacity factor

II Increasing unplanned forced outages

• Rising OM & A

• Increasing threat of institutional shut down

• Suggests that the systems run management;management are not running the systems

System Surveillance

(

system surveillance library/20051809.pdf · 2011. 9. 16. · {-j aspects monitored system...

Documents